Optimal Measurement Point Selection for Radio Environment Maps: A Two-Stage Approach

Electromagnetic (EM) situational awareness is essential for the deployment and operation of local and temporary non-public networks (NPNs). Radio Environment Maps (REMs) provide an effective basis for network planning and operation by capturing the prevailing EM conditions. However, efficient measurement location selection for REM construction remains a key challenge, particularly under strict time and cost constraints. This paper presents a two-stage measurement selection framework that employs Optimal Experimental Design (OED)-based strategies to generate a reduced candidate set during a pre-selection phase. In a subsequent optimization stage, combinatorial optimization is applied to select an improved subset of measurement locations by minimizing the Mean Absolute Error (MAE) of the resulting REM, using a scenario-specific propagation simulation model. The results demonstrate that the proposed framework reduces the number of required measurement locations while consistently improving REM construction accuracy across all evaluated strategies. Validation using real-world measurement data confirms the practical applicability of the approach, with the optimization stage achieving MAE reductions from about 4% to around 50%, independent of the chosen pre-selection strategy.

Multi-Agent Reinforcement Learning for Base Station Placement in TDOA-Based Localization

Accurate localization of devices is a key capability for emerging 5G and 6G networks and depends on effective base station (BS) placement. Conventional geometry-based approaches such as Geometric Dilution of Precision (GDOP) ignore realistic propagation effects such as Non-Line of Sight (NLOS) shadowing and multipath-induced Time of Arrival (TOA) bias caused by buildings. This paper proposes a ray-tracing-assisted Multi-Agent Reinforcement Learning (MARL) framework for environment-aware BS placement in Time Difference of Arrival (TDOA) localization systems. Proximal Policy Optimization (PPO) agents are trained on Channel Impulse Responses (CIRs) generated from a detailed 3D model of a university campus. Each agent cooperatively places one BS while optimizing a shared reward that combines localization accuracy and coverage. The approach is evaluated on five campus segments with varying propagation characteristics. Results show that the learned policy achieves localization accuracy comparable to conventional GDOP-based placement, lowering the average localization Mean Absolute Error (MAE) by about 3 % relative to the stronger (mean-optimized) geometric baseline. The behavior is segment-dependent, with a clear improvement on individual segments (up to about 14 %) and comparable or slightly higher error on the others. These findings indicate that incorporating site-specific propagation data into the placement process can match and selectively improve upon purely geometric strategies, motivating further work toward consistent gains.